Insufficient functional ?-cell mass can cause type 2 diabetes (T2D). This application will elucidate how the DNA methylomes in embryonic multipotent pancreatic progenitor cells (mPPCs) pre-determine this mass. During embryogenesis, mPPCs activate pro-endocrine transcription factor (TF) Ngn3 to give rise to endocrine progenitor cells (EPCs), from which ? cells are derived. DNA methylation is a relatively stable repressive mark that is laid down by several enzymes including DNA methyl-transferase 1 (Dnmt1). DNMT1 polymorphisms are associated with attenuated ?-cell function and human T2D. We show that fetal exposure to maternal low-protein diet, known to compromise functional ?-cell mass, enhances Dnmt1 expression in mouse mPPCs. We recently published that both mPPC and EPC pools can be split into subsets that carry different DNA methylation levels at likely instructive gene enhancers, which can bias islet-cell fate choice. Moreover, higher Dnmt1 expression in mPPCs/EPCs favors ?-cell differentiation; yet EPC differentiation toward islet cells involves a rapid and substantial downregulation of Dnmt1 and concordant de-methylation in putative enhancers of genes that regulate ?-cell proliferation and function (referred to as fitness). The latter includes Synaptotagmin 7 (Syt7), a gene that promotes glucose-stimulated insulin secretion. Our model is that the methylomes in mPPCs can influence the evolving methylomes in their descendant EPCs and ? cells. Thus, mPPCs with distinct methylomes can give rise to ?-cell subsets with different fitness by pre-setting the expression levels of genes that will be activated later for high ?-cell fitness qualities. Perturbing the early-stage methylomes, by changed Dnmt1 expression and/or maternal diet manipulation, will shift the portions of ?-cell subsets and consequently the functional ?-cell mass and the susceptibility to T2D. Consistent with this model, we showed that a portion (~55%) of ? cells derived from a subset of Ngn3+ EPCs that co-express TF Myt1 (i.e., Myt1+Ngn3+ or ?M+N+?) has higher postnatal fitness than those from M-N+ cells under normal physiology. Here, we will first define the fitness of M+N+ or M-N+ progenitor-derived ?-cell subsets under metabolic stresses such as aging, hyperglycemia, or insulin resistance to elucidate their physiological significance (Am 1). We will then examine the transcriptomes/methylomes of the two subsets to define the key methylated loci that can account for their distinct ?-cell fitness, including testing if the differential Syt7 expression in the two ?-cell subsets (observed by us) contributes to their different insulin secretion activities (Aim 2). Lastly, we will examine how maternal low-protein diet exposure impacts the derivation and fitness of M+N+ and M-N+ progenitor-derived ?-cell subsets, focusing on the roles of enhanced Dnmt1 expression in mPPCs (Aim 3). We expect to impact the field by establishing direct mechanistic links between maternal factors and genetic networks that regulate functional ?-cell mass and susceptibility to T2D. Our innovative combinatorial lineage marking of ?-cell subsets put us in a unique position to study these issues.